Updated golang version to 1.13.x and
also updated user to set GO111MODULE=on
and CGO_ENABLED=1 when doing development
in cephcsi
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
The problem happens when multiple PVCs with the
same UUID are attached/mounted on a node. This
can happen after creating a PVC from a snapshot,
or cloning a PVC.
make nouuid as the default mount option if
the format type is xfs to avoid mounting
issues.
updates: #966
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
The gocyclo linter complains about the high complexity of the
CreateVolume() function:
> pkg/rbd/controllerserver.go:133:1: cyclomatic complexity 21 of func `(*ControllerServer).CreateVolume` is high (> 20) (gocyclo)
By splitting it up and separeting the creation of an exisint CSI Volume
object in buildCreateVolumeResponse(), the gocyclic linter does not
complain any longer.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
golint has a pretty struct stylechek, it down not allow different
variable names for methods on an object:
pkg/rbd/rbd_util.go:970:1: receiver name rbdVol should be consistent with previous receiver name rv for rbdVolume (golint)
func (rbdVol *rbdVolume) ensureEncryptionMetadataSet(ctx context.Context) error {
^
pkg/rbd/rbd_journal.go:166:26: ST1016: methods on the same type should have the same receiver name (seen 2x "rbdVol", 3x "rv") (stylecheck)
func (rbdVol *rbdVolume) Exists(ctx context.Context) (bool, error) {
^
Rename the 'rbdVol' variable to 'rv' to make it consistent.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
rbdVolume.open() was split from commit 5dd34732e1e while moving part of
the functionality to util.ClusterConnection. It seems that .open() is
not used anywhere at the moment, so drop it until follow-up patches
require it again.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
The shared util.ClusterConnection can be used for rbd.rbdVolume and
cephfs.volumeOptions to connect to the Ceph cluster. This will then use
the shared ConnPool, and functions for obtaining connection details will
be the same across cephfs and rbd packages.
The ClusterConnection.Creds credentials are temporarily available until
all the functions have been adapted to use go-ceph and the connection
from the ConnPool.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
The name of the CephFS SubvolumeGroup for the CSI volumes was hardcoded to "csi". To make permission management in multi tenancy environments easier, this commit makes it possible to configure the CSI SubvolumeGroup.
related to #798 and #931
golint warns about the following statements:
ceph-csi/internal/util/csiconfig.go
Line 49: warning: exported function Mons should have comment or be unexported (golint)
ceph-csi/pkg/util/volid.go :
Line 72: warning: exported method CSIIdentifier.ComposeCSIID should have comment
or be unexported (golint)
Reported-by: https://goreportcard.com/report/github.com/ceph/ceph-csi
Updates: #975
Signed-off-by: Yug Gupta <ygupta@redhat.com>
The `deploy:` section is executed as part of each successful job. That
means the container images and Helm charts are built and pushed many
times during testing after a push/merge event.
By creating a deploy stage and listing it after the tests, it is
guaranteed to run only once after all tests have succeeded. By adding a
condition to only run the stage for push events, the images and Helm
charts should only get built+pushed after a PR has been merged.
See-also: https://docs.travis-ci.com/user/conditional-builds-stages-jobs#Specifying-conditions
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Instead of creating the tmp directory in
ceph-csi create temp directory outside
and use it to push helm charts
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
With extra logging, there is no need to call `travis_wait` anymore. In
addition the `travis_wait` command blocks output, so the build steps are
not reported until the script finishes.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
By repeated logging of the current step, Travis CI should receive some
terminal activity and is expected to not timeout anymore.
Fixes: #982
Signed-off-by: Niels de Vos <ndevos@redhat.com>
While introducing scripts/build_step.inc.sh the tests start to fail as
the script is included and each shell script is tested separately. By
adding the option --external-sources to shellcheck, the related warnings
are not reported.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
If Docker does not exist, use Podman.
Use all CPUs for the build, which hopefully will make it faster.
Signed-off-by: Yaniv Kaul <ykaul@redhat.com>
Fixes: #987
Running the containerized-tests is relatively time consuming as it runs
all the tests. By adding a TARGET=.. option to the Makefile, it is now
possible to run a selected test, for example:
make containerized-test TARGET=static-check
This reduces the time the tests take, so it is more friendly to the
developers while addressing problems.
When no TARGET=.. parameter is passed, the default tests are run.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
IneffAssign warns about the two following statements:
Line 1342: warning: ineffectual assignment to rFound (ineffassign)
Line 1350: warning: ineffectual assignment to zFound (ineffassign)
rFound and zFound should be set before entering the loop, otherwise the
initial value will overwrite the updated value on each iteration.
Reported-by: https://goreportcard.com/report/github.com/ceph/ceph-csi
Updates: #975
Signed-off-by: Niels de Vos <ndevos@redhat.com>
This new journal package isolates journal logic from the rest of util
and helps draw bright lines between what is a generic utility function
and what is csi journal logic.
Done partly as preparation for making use of go-ceph in journal.
No functional changes are made except to update references to allow the
code to compile.
Signed-off-by: John Mulligan <jmulligan@redhat.com>
The NewErrSnapNameConflict will allow packages outside of "util" to
create new instances of the ErrSnapNameConflict error.
Signed-off-by: John Mulligan <jmulligan@redhat.com>
This PR addes GRPC 1.27 to required section in go.mod
file. this is overridden by 1.26 in replace section
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
The internal/ directory in Go has a special meaning, and indicates that
those packages are not meant for external consumption. Ceph-CSI does
provide public APIs for other projects to consume. There is no plan to
keep the API of the internally used packages stable.
Closes: #903
Signed-off-by: Niels de Vos <ndevos@redhat.com>
create ns before the helm create to avoid
`Error: create: failed to create: namespaces
"xxx" not found` issue
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
By setting the WORKDIR in the container image, there is no need to pass
it on the commandline in the Makefile. This makes the line for the make
target a little cleaner.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
with the help of qemu-user-static we can run
different architecute contains.
more info at https://github.com/multiarch/qemu-user-static
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
Add support for multi architecture build
for cephcsi. with multistage docker build
we we build cephcsi binary for both arm64
and amd64 architecture.
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
This reduces the number of concurrent jobs that get executed. The "build
test" jobs are relatively quick, and give developers the 1st feedback.
Longer e2e testing now only happens once the build tests have passed.
See-also: https://docs.travis-ci.com/user/build-stages/
Signed-off-by: Niels de Vos <ndevos@redhat.com>
`go mod verify` does not fail when all moduled under vendor/ are
updated. However it does update `go.mod` in case it does not reflect the
downloaded versions.
By adding this additional verification in `make mod-check`, future
updates to vendor/ will require a correct go.mod as well.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
The recent update to k8s 1.18.0 causes an issue in go.mod that gets
corrected by running `go mod verify`. grpc should be used in version
1.26, and not in the expected update 1.27.
By removing the dependency and keeping grpc in the 'replace' section,
`go mod verify` seems to be happy.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Add the containerized test to the existing containerized-build job so
that resources are shared a little.
Both tests use a different distribution, similar to what a developer has
on their system, and production deployments on a server.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
`make containerized-test` has been added as a make target. This runs the
'make test' target in a container. All test dependencies are installed
in the container once, and the container is kept around for running
`make containerized-test` subsequently.
The test container is based on Fedora:latest so that all test tools get
easily installed and are available in a recent version. The production
container is based on the Ceph container which has CentOS as Operating
System and therefor a more stable (too old) toolset.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
